We have been invited by Masahiro Suzuki to give a presentation and discuss music transcription at Yamaha innovation road, Hamamatsu.
title of the talk:
Automated Music Transcription based on Formal Tree Language Models
abstract
We present some of our work on Automatic Musical Transcription, the problem of converting a musical performance into musical notation, and the tree-structured representation of musical scores that we use for this purpose. In order to extract a structured output from a sequence of timestamped input events (e.g. a MIDI file), we apply quantitative parsing techniques, based on prior hierarchical language models of musical notation, defined by generative weighted tree grammars.
We will see how this approach helps to make relevant and interdependent decisions and to calculate optimal transcription solutions with respect to multiple criteria such as the adequacy of the output to the input, and a measure of the readability of the notation. musical. The intermediate tree representation can be exported to XML-formatted score files (MEI), and some transcription case studies for monophonic and drum datasets will be presented.
